metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: rejected
dtype: string
- name: chosen
dtype: string
splits:
- name: train
num_bytes: 25491
num_examples: 99
- name: test
num_bytes: 1537
num_examples: 6
download_size: 20372
dataset_size: 27028
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*